Process Mining to Drive Transformation at Police Service
A UK-based police service wanted to improve response and case processing time for incidents. Together with BPM-D, they deployed process mining to analyse response data, identify bottlenecks/issues, and define potential initiatives which would improve how cases are managed.
The police service in focus had an internal team dedicated to delivering Operational Excellence who recognised the value of process management. There was an established process modelling & repository management platform and they had been investigating the use of Process Mining technology to supplement this process capability. There was scope for improvement in their handling of a specific category of incidents, with potential inefficiencies in both response and case processing. The police service acknowledged this and looked to deliver this improvement, leveraging their internal process capability alongside both Subject Matter Expert and BPM-D consulting support.
BPM-D’s Rapid Process Improvement (RPI) methodology was used to define how the handling of incidents should be transformed by using a 7-step approach.
Challenge & Opportunity
The Integrated Supply Chain function at the global technology company was implementing a new digital solution to assist with demand planning, supply planning, production planning, and the new S&OP processes.
With several ERPs and S&OP process maturity levels across business units from project-based items to physical containers to engine motors and parts, this company lagged capabilities to foresee Demand and effectively plan Supply and Production accordingly.
The implementation of the new digital solution to address the appropriate processes was estimated to save approximately $55M over five years, therefore fast and efficient actions were required.
This created a framework for prioritising improvements downstream and helped to focus analysis on the most critical areas of the process.
This police service uses a custom system for recording and managing incidents. This system creates event logs when used; these track what function the system has been used to perform and the time at which the function was executed. Further, rich data is also recorded which provides information such as the borough, user, and incident category, etc. The event logs and the rich attribute data for a 3-month period were extracted, anonymised, and then loaded into a process mining tool for analysis.
The subsequent analysis highlighted bottlenecks, trends, and potential areas for improvement. After validating the data interpretation with the System Owners, these insights were reviewed with process subject matter experts.
In the SME interviews, first the As-Is process was discussed to identify key, non-system driven tasks. Once there was agreement on the current processes, the data insights and SME pain points were discussed. This combination of quantitative and qualitative investigation generated a rich understanding of where transformation needed tobe delivered.
The As-Is landscape was validated through a workshop with a focused group of process stakeholders, fostering alignment and ensuring the interpretation was accurate.
The Process Mining Factor:
Generate an As-Is Process
- Process models were previously built through interviewing SMEs. This meant that they showed what an organisation thought it was doing. Process mining is built off of system extracts – the data does not lie. This builds a true representation of business processes, including repetition, failure points and bottlenecks.
Analysis by Segment
- Process mining investigations can be filtered by any attribute – case category, region/team, time of incident, etc. This enables the identification of issues and trends specific to a focused segment of the organisation, allowing for focused improvement and/or benchmarking.
- By using system data, you have visibility of how your organisation functions. With a continuous data feed, you can identify potential non-compliant behaviour and correct it before it is an issue. By analysing historical data, long term trends can be found and corrected.
A compelling, fully costed business case that yielded the following benefits: